Asset Audience Gap Recommendation and Insight

Abstract

Techniques for determining an audience gap for a communication campaign presented herein. The system can obtain data indicating a content item for a communication campaign of a client account. The system can determine, based on the attribute, that the content item is associated with a first group type. The first group type can include a plurality of audience segments. Additionally, the system can determine, using performance data of the communication campaign of the client account, a first performance value of a first audience segment in the plurality of audience segments. Moreover, the system can determine, based on the first performance value transcending a performance threshold value, that the content item has an audience gap associated with the first audience segment. Furthermore, the system can perform an action based on the determination that the content item has the audience gap associated with the first audience segment.

What is claimed is: 
     
         1 . A computer-implemented method, comprising:
 obtaining data indicating a content item for a communication campaign of a client account, the content item having an attribute;   determining, based on the attribute, that the content item is associated with a first group type, wherein the first group type comprises of a plurality of audience segments;   determining, using performance data of the communication campaign of the client account, a first performance value of a first audience segment in the plurality of audience segments;   determining, based on the first performance value transcending a performance threshold value, that the content item has an audience gap associated with the first audience segment; and   performing an action based on the determination that the content item has the audience gap associated with the first audience segment.   
     
     
         2 . The method of  claim 1 , the method further comprising:
 determining, based on performance data of a plurality of similar accounts, a set of features in content items that are presented to the first audience segment, wherein the set of features improves a performance metric of the content items that are presented to the first audience segment, and   wherein the action performed is further based on the set of features.   
     
     
         3 . The method of  claim 2 , wherein a similar account in the plurality of similar accounts has a similarity score that exceeds a similarity threshold value, the similarity score being derived based on a search query similarity between the client account and the similar account. 
     
     
         4 . The method of  claim 2 , wherein each similar account in plurality of similar accounts have a similar geolocation as the client account. 
     
     
         5 . The method of  claim 2 , wherein the client account includes a media asset profile, the method further comprising:
 comparing a first feature in the set of features and the media asset profile; and   generating, based on the comparing of the first feature and the media asset profile, a subset of features by removing the first feature from the set of features, and   wherein the action performed is based on the subset of features.   
     
     
         6 . The method of  claim 5 , wherein the comparing of the first feature and the media asset profile includes:
 processing the first feature, using a machine-learned model, to generate feature embeddings;   processing assets in the media asset profile, using the machine-learned model, to generate profile embeddings; and   comparing the feature embeddings and the profile embedding to determine that the first feature does not correspond to assets in the media asset profile.   
     
     
         7 . The method of  claim 1 , further comprising:
 processing a plurality of content items associated with the first group, using a machine-learned model, to determine the plurality of audience segments, a performance value of each audience segment in the plurality of audience segments exceeds a threshold value.   
     
     
         8 . The method of  claim 1 , the method further comprising:
 generating, using a machine-learned asset generation pipeline, a suggested asset based on the first audience segment, and   wherein the action performed includes a presentation of the suggested asset on a graphical user interface of the client account.   
     
     
         9 . The method of  claim 8 , wherein the client account includes a media asset profile, and wherein the suggested asset is generated using the machine-learned asset generation pipeline further based on the media asset profile of the client account. 
     
     
         10 . The method of  claim 8 , wherein the suggested asset is generated based on a pre-existing asset associated with the client account, the pre-existing asset having been previously uploaded to the client account. 
     
     
         11 . The method of  claim 8 , the method further comprising:
 processing the suggested asset and the first audience segment, using a machine-learned estimation model, to determine an uplift estimation value, and   wherein the action performed further includes the presentation of the uplift estimation value on the graphical user interface.   
     
     
         12 . The method of  claim 11 , wherein the machine-learned estimation model being trained on performance data of previously presented content items of another account. 
     
     
         13 . The method of  claim 1 , wherein the first performance value is a click-through rate, and the threshold value is an average click-through rate for the first audience segment. 
     
     
         14 . The method of  claim 1 , wherein the first group being a type of product, and wherein the content item is currently being served in the communication campaign. 
     
     
         15 . The method of  claim 1 , wherein the plurality of audience segments is determined based on a conversion rate for content items that are associated with the attribute. 
     
     
         16 . The method of  claim 1 , the method further comprising:
 determining the plurality of audience segments for the communication campaign based on the attribute of the content item, wherein each audience segment in the plurality of audience segments is associated with the attribute.   
     
     
         17 . The method of  claim 1 , the method further comprising:
 obtaining past performance data of content items that are associated with the first group type, and   wherein the performance threshold value is calculated based on the past performance data of content items that are associated with the first group type.   
     
     
         18 . The method of  claim 1 , the method further comprising:
 obtaining past performance data of content items that are associated with a plurality of accounts that are similar to the client account, and   wherein the performance threshold value is calculated based on the past performance data of content items that are associated with the plurality of account that are similar to the client account.   
     
     
         19 . The method of  claim 1 , the method further comprising:
 obtaining a plurality of content items that are associated with the first group type; and   determining, based the first audience segment, a set of features in content items that are presented to the first audience segment, wherein the set of features improves a performance metric of the content items that are presented to the first audience segment; and   generating, based the set of features and the plurality of content items that are associated with the first group type, a subset of features by removing a first feature from the set of features, and wherein the plurality of content items do not include the first feature.   
     
     
         20 . A computing system comprising:
 one or more processors; and   one or more transitory or non-transitory computer-readable media storing instructions that are executable to cause the one or more processors to perform operations, the operations comprising:
 obtaining data indicating a content item for a communication campaign of a client account, the content item having an attribute; 
 determining, based on the attribute, that the content item is associated with a first group type, wherein the first group type comprises of a plurality of audience segments; 
 determining, using performance data of the communication campaign of the client account, a first performance value of a first audience segment in the plurality of audience segments; 
 determining, based on the first performance value transcending a performance threshold value, that the content item has an audience gap associated with the first audience segment; and 
 performing an action based on the determination that the content item has the audience gap associated with the first audience segment. 
   
     
     
         21 . One or more non-transitory, computer readable media storing instructions that are executable by one or more processors to cause a computing system to perform operations, the operations comprising:
 obtaining data indicating a content item for a communication campaign of a client account, the content item having an attribute;   determining, based on the attribute, that the content item is associated with a first group type, wherein the first group type comprises of a plurality of audience segments;   determining, using performance data of the communication campaign of the client account, a first performance value of a first audience segment in the plurality of audience segments;   determining, based on the first performance value transcending a performance threshold value, that the content item has an audience gap associated with the first audience segment; and   performing an action based on the determination that the content item has the audience gap associated with the first audience segment.